Update app.py

app.py

@@ -33,7 +33,7 @@ tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
 # Load the Lora model
 model = PeftModel.from_pretrained(model, peft_model_id)
 
-@spaces.GPU(duration=200)
+@spaces.GPU()
 def get_completion(query, model, tokenizer):
     try:
         # Move model to CUDA
@@ -56,7 +56,7 @@ def get_completion(query, model, tokenizer):
         model = model.cpu()
         torch.cuda.empty_cache()
 
-@spaces.GPU(duration=200)
+@spaces.GPU()
 def code_review(code_to_analyze):
     query = f"As a code review expert, examine the following code for potential security flaws and provide guidance on secure coding practices:\n{code_to_analyze}"
     result = get_completion(query, model, tokenizer)